IMRT for H&N Cancer

K.S. Clifford Chao, M.D.

Associate Professor Director of Molecular Image-guided Therapy Department of Radiation Oncology University of Texas, M.D. Anderson Cancer Center Houston, Texas

Outline
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study

Outline
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study

Fabrication of Non-invasive Immobilization Thermoplastic Mask for IMRT

CT Simulation for H&N IMRT

Step-n-Shoot IMRT by Segmental MLC

QA for Step-n-Shoot IMRT by Segmental MLC

Outline
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study

Target Determination

What to Shoot?

Clinical Target Volume Determination for Head and Neck IMRT

Tumor Site
Oral Cavity Buccal RMT Oral Tongue
T1-2NO T3-4N+* N2c T1-2NO T3-4N+* N2c T1-2NO T3-4N+* N2c T1-2NO T3-4N+* N2c GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n IN (I-III) IN+CN(I-V) IN (I-IV) IN+CN(I-V) IN (I-III) IN+CN(I-V) IN (I-III) CN (I-III) IN +/- CN (I-IV) CN (I-IV) IN (I-III) CN (I-III)

Clinical Presentation

CTV1 (70Gy/35fx)

CTV2 (63Gy/35fx)

CTV3 (56Gy/35fx)

FOM

Nasopharynx

IN (I-V, RPLN) IN+CN (I-V, RPLN)

IN+CN (I-V, RPLN) CN (I-V, RPLN)

Chao et al., IJROBP 53:1174, 2002

Clinical Target Volume Determination for Head and Neck IMRT

Tumor Site
Oropharynx BOT
T1T1-2NO T3T3-4N+* N2c T1T1-2NO T3T3-4N+* N2c T1T1-2NO T3T3-4N+* N2c T1T1-2NO T3T3-4N+* N2c GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n GTVp GTVp+n GTVp+n IN (I(I-V, RPLN) IN+CN (II(II-IV, RPLN) IN (I(I-V, RPLN) IN+CN (II(II-IV, RPLN) IN (II(II-IV, RPLN) IN+CN (I(I-V, RPLN) IN (II(II-V) IN+CN (I(I-V) IN+CN (II(II-IV, RPLN) CN (I(I-V, RPLN) IN+/IN+/-CN (II(II-IV, RPLN) CN (I(I-V, RPLN) IN+CN (II(II-IV, RPLN) CN (II(II-IV, RPLN) IN+CN (II(II-V) CN (II(II-V)

Clinical Presentation

CTV1 (70Gy/35fx)

CTV2 (63Gy/35fx)

CTV3 (56Gy/35fx)

Target Delineation

Tonsil

Hypopharynx

Larynx

Chao et al., IJROBP 53:1174, 2002

Nodal Target for Head and Neck IMRT

For N0 neck, which one will you pick?

Target Delineation of Clinically N+/N- Necks in Patients Receiving Definitive IMRT

CTV1

CTV2
P

III

CTV1

CTV2

NR

II

Wijers, 1999

Gregoire, 2000

Nowak, 1999

GTV

N+

N-

N+

N-

Chao et al., IJROBP 53:1174, 2002

Incidence of Extracapsular Extension of Metastatic Neck Node by Size

Nodal size Annyas 1979 Johnson 1981 Carter 1987 Hirabayashi 1991 <1cm 23% 17% 43% 1-3cm 53% 65% 83% >3cm 74%

Target Delineation of Clinically N+/N- Necks in Patients Receiving Definitive IMRT

CTV1 CTV2 GTV

CTV2

CTV1
Ib

75% 95% 81%

GTV

II

II

N-

N+

N-

N+

Chao et al., IJROBP 53:1174, 2002

Target Delineation of Pathologically ECE+/ECE- Necks in Patients Receiving Post-operative IMRT

CTV1

CTV2

CTV1

CTV2

Examples

ECE+

ECEECE-

Chao et al., IJROBP 53:1174, 2002

T2N2bM0 SCC of Tonsillar Fossa

IMRT Target Dose Specification

Target Volume
Concurrent Chemotherapy

H&N IMRT
Butler NO
All Sites

RTOG H-0022

Lee Yes
NPC

Chao Yes
All Sites

70Gy/35fx

NO
Early Oropharynx

CTV1
70Gy/35fx

60 / 2.4Gy 25fx 50 / 2Gy

66 / 2.2Gy 30fx 60 / 2Gy 54 / 1.8Gy

70 / 2.12Gy 33fx 59.4/1.8Gy -

70 / 2Gy 35fx 63 / 1.8Gy 56 / 1.6Gy

56Gy/35fx

CTV2 63Gy/35fx
60Gy/30fx

CTV3
50Gy/25fx

T2N1M0 SCC of Base of the Tongue

70Gy/35fx

Target Delineation of BOT Carcinoma

63Gy/35fx

56Gy/35fx

Post-operative IMRT

Sensitivity and Specificity of CT and MRI in Detecting Clinically Negative but Pathologically Positive Neck Nodes
Author Stern 1990 Friedman 1990 Moreau 1990 Van den Brekel 1993 Righi 1997 Modality CT CT MRI CT CT MRI CT Pt. No. 53 68 16 32 49 55 32 Sensitivity 40% 68% 80% 50% 78% 88% 50% Specificity 92% 90% 82% 86% 86% 83% 86%

MR Fusion for NPC Target Delineation

MR Fusion for NPC Target Delineation

MR Fusion for NPC Target Delineation

Target Delineation of Critical Structures

IMRT is quid pro quo something for something

GTV

320

40

260 230 200 160

120

HOE

HOE
310 260 230 120 200 180 160 50

After

Before

More Efficient??
Class Solution Sharing Experience

www.imrttarget.org

Outline
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study

Therapeutic Outcome of Oropharyngeal Carcinoma

Washington University Experience (1970-1999)

Patient No. Def. CRT Def. IMRT PostPost-op CRT PostPost-op IMRT 153 12 142 14

Median F/U 3.5 yr (1.6(1.6-17.7) 2 yr (1(1-3) 3.9 yr (1.3(1.3-19.8) 2.2 yr (1(1-3.2)

2yr LC 68.3% 87.5% 75.7% 100%

2yr DFS 58.4% 79.5% 73.5% 92.5%

Chao et al. Radiotherapy & Oncology, 2001

Therapeutic Outcome of Oropharyngeal Carcinoma

Washington University Experience (1970-1999) Def. CRT (n=153) Def. IMRT Post-op CRT Post-op IMRT (n=12) (n=142) (n=14)

Therapeutic Outcomes of Published Head and Neck IMRT Series

Author IMRT Planning N Subsite LC (%) Butler Inverse planning Forward planning 20 Multiple 58 Multiple N/A N/A LRC (%) 85* 79 (2-y) 75 (5-y) OS (%) N/A N/A

Acute Grade 3-4 mucositis Late Grade 2-3 xerostomia

(12m post-RT)

25%
P=0.134

42%

20%
NS

21%

Dawson

84%
P<0.001

30%

77%
P<0.001

17%

Lee Chao

Inverse and forward Inverse planning

67 NPC

97 (4-y) 98 (4-y) 88 (4-y)

126 Multiple 92 (3-y) 83 (3-y) 85% (3-y)

Chao et al. Radiotherapy & Oncology, 2001

Materials and Methods

Chao et al. IJROBP 2003

Patterns of Failure

From 2/97 to 12/00 126 head and neck patients (96 male, 30 female) Median age 56 (range 1313-84 years) 52 definitive, 74 postop 35 definitive IMRT patients received chemotherapy Median followfollow-up 29 months (range 1919-62)

Materials and Methods

Primary site
Oropharynx 63 Oral cavity 15 NPX 12 Para. &nasal cav. cav. 9 Unknown primary 9 Larynx 7 HPX 8 Other sites 3 Def IMRT 72.644.83 Gy to CTV1 64.345.15 Gy to CTV2 Postop IMRT 68.534.71 Gy toCTV1 60.955.33 Gy to CTV2

Head and Neck IMRTIMRT-Results

Overall Def. IMRT 80 85 85 Postop IMRT 90 94 88 P value p=0.36 p=0.42 p=0.58 2-year LRCR 2 year ult. ult. LRCR 2-year OS 86 90 87

T stages
T1 T2 T3 T4 N stages N0 N1 N2 N3 AJCC staging Stage I Stage II Stage III Stage IV 19 33 27 38 30 26 61 9

1.0

Postop IMRT Def IMRT

.8

Disease Free Survival

5 8 26 78

.6

.4

P=0.009

.2

0.0 0 10 20 30 40 50 60 70

Months

Geographic Detail of LR Failures

No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Tumor Site L. Tonsil R. PS BOT L. Tonsil R. Tonsil L. Tonsil L. RMT NPX L. Tongue Tonsil R. Tongue L. PS L. BOT L. larynx L. RMT UP L. Tongue T 4 4 4 4 3 2 3 3 4 2 2 3 3 3 4 X 2 N 2A 1 2C 3 2B 2A 0 3 1 2A 2 0 2C 2B 2B 2C 0 AJCC IV IV IV IV IV IV III IV IV IV IV IV IV IV IV II Aim Def Def Def Def Def Def Def Def Def Def Postop Postop Postop Postop Postop Postop Postop Site of Failure L. Tonsil R. Level II LN Bilateral neck R. Level II LN R. Tonsil L. Tonsil L.RMT Lower neck L. Tongue, neck Lower neck R. Level III LN L. Level II LN R. Level IB LN L. Level II LN Thyroid, Level IA LN R. Neck L. level III/IV LN Location CTV1 CTV1 CTV1 CTV1 CTV1 CTV1 CTV1 Lower neck CTV1 Lower neck Lower neck CTV2 CTV2 CTV2 CTV1 Lower neck Post. Neck Lower neck Time to rec. (months) 11 17 10 Pers. 10 Pers. Pers. 12 12 10 9 6 2 12 6 7 8

Pre-RT MRI
Tumor

Post-RT MRI

Example 1: FC
Persistent tumor

Persistent tumor

CTV1 CTV2

Persistent tumor

Chao et al. IJROBP 2003

Chao et al. IJROBP 2003

What Have We Learned?

We observed no parotid or dermal failure. Satisfactory local-regional control in CTV2. Local failure predominantly within high dose regions Need to discern radioresistant subpopulation within CTV1

Outline
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study

Biological Modifiers

Marginal Failure or In-field Failure?

Marginal Failure In-field Failure

CTV2

CTV1

Increase Volume Decrease Volume

Volume

Dose Target

Dose Escalation Dose Reduction

Better Target Delineation Biological Causes

Refine Target Volume Bigger or Smaller? NN+

CT-FDG PET Imaging Co-registration for IMRT Target Delineation A

CT-FDG PET Imaging Co-registration for IMRT Target Delineation

Marginal Failure or In-field Failure?

Marginal Failure In-field Failure

Imaging Targets of Tumor GenoGenoand PhenoPheno-type

Gene expression Receptor, kinase

CTV2

CTV1

Better Target Delineation

Biological Causes

DNA Synthesis Energy Metabolism

N-

N+

Blood Flow and Hypoxia

Tumor Hypoxia Contributes to Local Failure?

Delineation of Hypoxic GTV by 60Cu-ATSM

Chao, IJROBP 2001; 49(4): 1171-1182

Problem (Example 1)

A Target Coverage Scoring Function for IMRT Planning

-Based on the Probability of Gross Disease, Microscopic Tumor Extension and Lymph Node Metastasis

Lack of Spatial Information in DVHDVH-based IMRT Optimization Algorithms

K.S. Clifford Chao, M.D.1 Angel I. Blanco, M.D. 2 James F. Dempsey, Ph.D.3
1M.D.

Anderson Cancer Center University Medical School of Florida

2Washington

3University

Plan 1

Plan 2

A Score Function to Discriminate Probability of Microscopic Extension from Gross Tumor & Lymph Node Metastasis for Voxels within CTV

Probability of Tumor Extension vs. Distance

10
0

Proba bility of Mic ro s copic Exte ns ion

10

-1

Breast Cancer l = -1.02 r = 0.98

10
-2

10

-3

Lung: Adenocarcinoma l = --5.14 r = 0.98

10

-4

Lung: Squamous C ell Carcinoma l = -6.77 r = 0.96

10

-5

Distance from GTV Surfac e (cm)

Probability of Nodal Metastasis vs. Distance in H&N Cancers

Ca uda l Sp re ad - Clinica lly LN Neg ative 1 Pro ba bility o f LN Me ta s tas is 0.8 0.6 0.4 0.2 0 Pro ba bility o f LN Me ta s tas is 1 Ce pha lic Sp re ad - Clinica lly LN Neg a tive

m = -0.021 r = 0.74

0.8 0.6 0.4 0.2 0

m = -0.033 r = 0.71

Score Function of Gross Disease, Microscopic Extension, and Nodal Metastasis

2 4 6 8 Dis ta nce from GTV (cm) Ca ud al Sp re ad - Clinica lly LN P os itive

10

2 4 6 8 Dis ta nc e from GTV (c m) Ce pha lic Spre a d - Clinica lly LN Pos itive

10

Gross Tumors

1 Pro ba bility o f LN Me ta s tas is 0.8 0.6 0.4 0.2 0 Pro ba bility o f LN Me ta s tas is

m = -0.065 r = 0.87

0.8 0.6 0.4 0.2 0

m = -0.12 r = 0.85

Small, but not zero

2 4 6 8 Dis ta nce from GTV (cm)

10

2 4 6 8 Dis ta nc e from GTV (c m)

10

Problem (Example 2)

Lack of Spatial Information in DVHDVH-based Optimization Algorithms

Gross tumor

Plan 3

Significantly Sparing Left Parotid

Plan 3

Metastatic Node

Plan 4

Plan 4
Gross tumor Significantly Sparing Right Parotid Metastatic Node

Solution #1
2-D Histograms for DoseDose-VolumeVolume-Tumor Extension/Metastasis Probability by Scoring Function

Plan 3

Underdosed volume located in the lower risk regions

Solution #1
2-D Histograms for DoseDose-VolumeVolume-Tumor Extension/Metastasis Probability by Scoring Function

Solution #2

Relative Residual Tumor Burden: Voxel (RRTBijk) & Integral (IRRTB)

nf SF ijk Model e = ijk RRTBijk = D 2 SF ijk DRx DRx Rx nf e Dijk Dijk 2

Plan 4

IRRTB =

SF SF
ijk ijk

ijk

Model

ijk

DRx

Underdosed volume located in the higher risk regions

Where RRTBijk is the Relative Residual Tumor Burden of the i-j-k-th voxel, SFModel is the surviving fraction of the model, SFDRx is the prescribed surviving fraction, and are radiosensitivity parameters ( = 0.29 Gy-1 and = 0.029 Gy-2 ), Dijk is the total plan dose of the i-j-k-th voxel, DRx is the total prescription dose, rijk is the scoring function value for the i-j-k-th voxel, and nf is the number of fractions.

Summary
Treatment and QA Techniques Target Determination and Delineation Clinical Results Ongoing Study